Risk & Compliance Project Management Data & Analytics
Consulting Pharma & Life Sciences Energy & Resources Manufacturing & Automotive Financial Services Consumer & Retail Technology & IT
By outcome
Early Warning Risk Visibility Predictive Intelligence Time Savings Financial Protection Institutional Memory
By use case
Budget Overrun PMO Reporting Portfolio Risk Visibility See all 10 use cases →
How We Work Technology Pricing Start Free Maturity Check →
Research

Monte Carlo Simulation in Investment Risk Analysis: What a €51 Million Project Teaches Us About Budget and Time Estimates

Platon & Constantinescu (2014) analyze a Romanian environmental project with 1,000 Monte Carlo simulations. The results are striking — and have direct consequences for your portfolio.

Platon & Constantinescu (2014)Procedia Economics and FinanceReading time: 5 min

Context: Why this paper matters

Since 1944, when Stanislaw Ulam and Nicholas Metropolis developed the Monte Carlo method at the Manhattan Project, it has been considered the gold standard for quantitative risk analysis. Yet most companies still use it for individual projects, if at all. Victor Platon and Andreea Constantinescu of the Romanian Institute of National Economy show in their 2014 study published in Procedia Economics and Finance how the method works in practice — and where its limits lie.

The project under study: An integrated waste management system in Suceava, Romania. Value: €51.76 million. Planned duration: 45 months. The question: How likely is it that the budget and schedule will be met?

Key findings

The authors divided the project into six cost components (tradable goods, non-tradable goods, skilled labor, unskilled labor, land acquisition, transfer payments) and four construction phases. For each component they defined a triangular distribution with minimum, maximum, and most likely values. Then they ran 1,000 simulations.

The numbers at a glance

Project value (estimated)€51.76M
Average from 1,000 sim.€50.75M
Probability under budget67.3%
Probability over budget32.7%
Project duration (estimated)45 months
Average from 1,000 sim.49.9 months
Probability under schedule4.8%
Probability over schedule95.2%
Risk of exceeding both31.1%

The result is a classic pattern we observe in hundreds of portfolios: Budgets are more pessimistic than necessary; schedules are more optimistic than realistic. The estimated project value was slightly above the simulation average — suggesting the cost estimate was conservative. The project duration, however, was almost five months below the simulation average — a dramatic difference that almost certainly leads to delays.

Critique: What the study does not address

The study is methodologically sound, but not without limitations:

The triangular distribution: A hidden problem

Platon and Constantinescu use a triangular distribution — the simplest probability distribution with only three parameters. The problem: it overemphasizes extreme values. In practice, project costs tend to follow normal or log-normal distributions. Choosing a triangular distribution systematically underestimates the probability of moderate deviations and overestimates the probability of extreme outliers.

Recommendation: What this means for your portfolio

The study's results can be distilled into three principles that any portfolio team can apply — regardless of industry or project size:

1. Budget and schedule are different risk categories

The study shows that budget and schedule estimates should not be lumped together. A project can be under budget and over schedule — or vice versa. An aggregated “project score” obscures these differences. Separate both dimensions in your reporting.

2. Real-time data beats manual estimates

The subjective parameters of the study are its biggest weakness. If your Monte Carlo simulation is based on estimates rather than real-time data, the output is only as good as your worst assumption. Connect your risk model to ERP, Jira, SAP, or SharePoint — so budget consumption, milestone progress, and resource utilization flow in automatically.

3. Simulate continuously, not once

A simulation at project start is better than none. But it becomes outdated quickly. Every new data point — a booked expense, a shifted milestone, a new risk — changes the probability distribution. If you simulate your portfolio only quarterly, you miss early warning signals for three months.

Aversight in context of the study

Aversight addresses exactly the three weaknesses of the study: real-time data from up to 15 connected systems replaces manual estimates. 2,500 iterations per project instead of 1,000. And instead of a static analysis, Aversight updates the probability bands hourly — with automatic alerts when a score threshold is exceeded. The result: no more single-project black boxes, but a transparent portfolio risk picture that sharpens with every new data point.

Conclusion

Platon and Constantinescu provide valuable evidence that Monte Carlo simulations in investment risk analysis are not just theoretically powerful, but practically implementable. The central insight — that schedules are systematically underestimated and budgets overestimated — confirms itself in almost every portfolio we analyze.

But the study's limitations (subjective parameters, static analysis, low iteration count) also show where the next step lies: from manual, one-time simulation to automated, continuous risk intelligence. Those who take this step shift risk management from reaction to prevention.

Risk intelligence is not a black box. Let us show you how it works.

30 seconds — and we will get back to you within 24 hours.

Start Free Maturity Check →